2
\$\begingroup\$

I created a simple power supply as shown below that results in an unregulated 50VDC which is what I desire. The transformer is rated for 1000VA. It works ... but only if you're lucky enough to turn it on at the right time of the AC cycle to not pop the breaker. Clearly I need to limit the inrush current on those capacitors.

I was looking at putting an NTC Termistor in series, probably between the transformer secondary and the bridge rectifier. Specifically SL22 5R012 which is 5 ohms at room temperature, can do 12A at steady state, and can survive 100J.

The problem I foresee however is the load connected to this supply is not constant. The load is DC solenoids that could not be activated for a significant amount of time (minutes to hours). That would mean the NTC would likely cool down correct? I don't know if the capacitors would be enough for the NTC to reheat when demand is needed. There would never be a constant load on the supply, aside from the bleeder resistors which is pretty low.

I've heard of having a relay or MOSFET to bypass an NTC or regular resistor after X seconds, but unsure how to implement it with only the 50V available.

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$
  • \$\begingroup\$ Bypassing the NTC with a relay once booted is always an option. \$\endgroup\$ – winny May 10 '18 at 12:53
2
\$\begingroup\$

You can try something like this. With the supply off, the limiting resistor is in the primary circuit. When the DC output reaches a level that can energize the relay, the resistor will be shunted.

As @Autistic mentioned, transformers can create surge currents when power is applied. Worst case is when the applied AC voltage happens to be at one of its positive or negative peaks. This can momentarily saturate the core. An interposing solid state relay that switches at zero crossing would prevent this, but you still have the filter capacitance initial charging current reflected back to the primary. The series limiting resistor addresses both sources of inrush.

I would choose a resistor power rating of 10 ~ 25 Watts.

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$
  • \$\begingroup\$ That seems reasonable. Wouldn't a fly back diode be needed on the relay? \$\endgroup\$ – Geomancer May 10 '18 at 16:52
  • \$\begingroup\$ @Geomancer Across the contacts? Probably not necessary, as long as the relay contacts are rated for AC at the current you need. That would require a bi-directional diode anyway (tranzorb) or R-C snubber. The relay will de-energize some time after AC power is removed, so its contact is dry when it opens. It closes with steady state current already passing through the resistor, so minimal transient effect. Diode across the coil? Not needed, since there is nothing sensitive like a MOSFET switching the coil off. The filter caps act as huge snubbers to the little energy developed by the coil. \$\endgroup\$ – AlmostDone May 10 '18 at 18:09
0
\$\begingroup\$

Transformers also do have inrush current .Large transformers are worse .Toroidal designs are worse than old school laminated types.You should place your proposed inrush current limiting in the primary circuit so the transformer and the filter caps are covered.Your proposed NTC approach is simple and better than nothing .If you bypass the NTC with a relay when the DC volts have settled then you will not waste power and will not end up with a hot NTC that will have a low resistance and not be effective.The primary NTC should be a lot more than 5 ohms in order to keep peak surge currents acceptable .

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.